AI Infra agents

一个利用pydantic ai实现的AIAgent，使其能在MCBE中使用Websocket服务器实现的游戏内调用LLM"In MCBE, a WebSocket server is used to enable the use of LLM within the game."

MCP server for nodiom — structured read/write access to Markdown documents for AI agents.

Companion OpenClaw skill for the Unreal plugin — teaches AI agents the tool surface and editor workflows so they can drive Unreal Engine remotely.

Companion OpenClaw skill for the Godot plugin — teaches AI agents the tool surface and editor workflows so they can drive Godot remotely.

轻量级跨平台 LLM 客户端 + AI Agent 开发环境 | Lightweight cross-platform LLM client & agent workspace (Tauri 2 + Vue 3 + Rust)

Fast, deterministic CLI to search and navigate Open Knowledge Format (OKF), or Obsidian style doc bundles — for humans and AI agents. Find by frontmatter, traverse the link graph, surface orphans/deadlinks.

AI-native software delivery governance framework

Persistent, structured memory for AI agents — a human-readable Markdown wiki, multi-user with per-fragment ACL, self-organizing overnight. Agent-agnostic via MCP. Self-hosted, AGPL.

Python SDK for BlockRun — every frontier model, pay-per-call with USDC. OpenAI-compatible. Zero rate limits.

MCP server that lets one AI agent ask other agents for help

Pattern for emulating named custom subagents with the Amp SDK: dynamically defined roles with their own system prompts, scoped tool permissions, and isolated context per invocation.

Short-term memory proxy gateway with proactive memory surfacing for AI agents

A board of expert personas whose every decision is a pre-registered, time-anchored, reality-graded bet. BYOK; the board that keeps score, before the fact.

A curated list of SOUL.md files for AI agents, personas, and characters.

Single-user, local agentic harness

Generative UI for AI agents on Microsoft Foundry: a .NET 10 Agent Framework agent over AG-UI, a CopilotKit runtime, and a React frontend on Azure. Charts, shared state, and human-in-the-loop.

An operating system for product managers — persistent product context, an on-demand skills library, and a continual-learning loop for AI-native PM work.

Run an AI company. Not a prompt. Open-source, privacy-first desktop app for AI-agent organizations — Windows, macOS, Linux.

A spec-driven pipeline that makes building with AI fast, accountable, and cumulative.

A community-curated list of MCP (Model Context Protocol) servers. PRs welcome! Browse AI tools, integrations, and developer resources.

基于 AI 的股票新闻分析系统，将新闻与股价走势关联，实现智能选股决策支持

Build UI apps with your AI assistant — together, live. An agent-native, cross-platform declarative-UI runtime in Go: describe UI as JSON, let your AI scaffold/edit/verify it over MCP, package for web/iOS/Android/desktop/mini-program.

Agent reasoning loop with tool calling for Gemma 4 models. Handles prompt construction, tool call parsing, execution, and multi-turn conversation management.

Open-source runtime for AI agents and multi-agent graphs — durable, declarative, self hosted.

AgentClash —— 面向 LLM Agent 的半回合制策略对战平台：写策略代码、云端模拟、ELO 排位

ИИ-ассистент для 1С:Предприятия. Расширение конфигурации (CFE) с чат-интерфейсом, on-premise Go-шлюзом, поддержкой LLM (OpenAI, Ollama) и MCP.

Cognitive runtime layer for AI engineering. Adds planning, memory hierarchy, and deterministic execution on top of any language model.

Pre-domain framework for transition dynamics and structural enforcement

This repository is based on a study that if the critic overcome the generators failures or they inherit them.

A lightweight, mobile-first Vue 3 chat UI component library built with Vant 4. Features 20+ customizable components including message bubbles, typing indicators, emoji picker, and rich text editor. Perfect for building AI chatbots, LLM interfaces, instant messaging apps, and customer service chat systems. Supports auto-scroll, typewriter effects.

telegram AI that runs on your account. Answers in groups and PMs from personal account

🤖 A web UI for large language models.

A modern, fully-async framework for Bale (بله) bots — inspired by aiogram.

These are the Microsoft Semantic Kernel Workshop notebooks. It addresses AI agents, agents collaboration and, of course, kernel, plugins, planners and function calling

Portable multi-tenant BYOK model registry — auto-updating models.dev catalog plus per-tenant credential resolution behind one resolveModel() call, with host-gated keys. Zero runtime dependencies, ESM, Node ≥22.6. Distilled from openclaw + hermes-agent.

[WIP] Sorai is an open-source LLM proxy gateway built for resource efficiency and high performance.

A semantic type system for validating LLM output intent — Pydantic for meaning.

RAG answers you can verify — citations, faithfulness checks, exportable evidence reports

CI-friendly RAG evaluation: retrieval metrics, LLM-as-judge faithfulness, and a regression gate that fails your build when quality drops

AI-powered educational assessment generation platform for instructors, trainers, and enterprise knowledge management systems.

agent-skills for chonkie's ecosystem

NotebookLab | Your thinking partner, on your machine. Import your documents. Ask questions. Get answers with sources. Write connected notes, search everything, and let a local AI help you think. Nothing leaves your computer.

Native macOS agent harness with permissioned tools, local LLM support, real document generation, and inspectable Markdown memory.

基于 Dify + Ollama + MCP 实现的全本地化电商客服智能体

One agent SDK, hand-ported to 7 languages — JavaScript, Python, Go, Java, C#, Elixir, Clojure — every port held to the same spec by a CI-enforced cross-port conformance suite, all seven at full tier. MCP · skills · native · HTTP · built-in · A2A tools behind one interface. Docs: https://muthuishere.github.io/toolnexus/

LangGraph and LangChain retriever, tools, and navigator subgraph for OKF (Open Knowledge Format) bundles.

Eticas AI Risk Taxonomy: Open, machine-readable AI risk taxonomy mapped to the EU AI Act, NIST AI RMF, ISO/IEC 42001, and 15+ other frameworks.

Penalize the Path, Reward the Outcome — verifiable per-action penalties as a dense channel for deployable, sample-efficient agentic RL (GRPO). Paper: arXiv:2607.07435

Developer infrastructure for discovering and comparing LLM models and providers

Query-aware context trimming for LLM requests. Your context could use a trim.

LES: local and distributed evidence system for project documents, verifiable retrieval and model tools.

Free offline AI with complete Wikipedia knowledge - 100% private

A model-facing field manual for vehicle reality, machine physics, lifecycle economics, failure diagnosis, and practitioner judgment.

Rust based JSON repair library

A multi-platform chatbot framework for LLMs. Connect to various large language models via multiple instant messaging apps — QQ, WeChat, Telegram, Discord, and more. 😄 Extensible with plugins and modules. 👍 Built on Python 3.11+ with a Quart-based web panel. Beginner-friendly — learn plugin development quickly. 🎉

Tools Are Not Islands: Set-Level Tool Retrieval for LLM Agents via Query-Conditioned Hyperedge Prediction

「萦回」AI互动叙事平台 - AI 互动叙事引擎——多 Agent 自主演化沙盒。创建架空世界，角色自主互动、冲突、和解，世界自行演化。支持记忆系统、安全风控、一键小说生成。

Not a chatbot. Not a coding assistant. A digital operator that controls your entire PC — 219 tools, screen perception, HID control, browser automation, Office suite, AI media. One sentence in, finished work out.

A free, private, local-first AI assistant that actually does the work — 167 tools, a team of 78 AI agents, 40 rotating free AI brains, 25 skills. Not a chatbot, a doer. It doesn't just answer — it handles it.

Light-weight, query-aware compression for high-signal LLM prompts.

A Godot-native framework for LLM-driven NPCs with tools, memory, and goal-oriented behavior

peft safetensors text-generation conversational en arxiv:2409.11500

Agentic task prioritization on Google's Agent Development Kit (ADK): agents that rank work against OKRs, explain their reasoning, and ship as a Dockerized service.

Real, reproducible LLM inference benchmarks on the NVIDIA DGX Spark — parallel agent sessions, tok/s and latency under load, with an interactive dashboard and one-command recipes

Vendor-neutral graph orchestration for reliable agentic systems: one portable Graph IR, native TypeScript/Python runtimes, deterministic barriers/routers, and cross-language conformance.

Context engineering decides what a model sees. 14 self-contained examples that build the mental model one piece at a time, Node.js + local GGUF. Understand deeply, then use frameworks wisely. Hands-on context engineering - all local.

A multi-agent AI system with isolated agent memory, deterministic confidentiality guards, transactional context compaction, provider adapters, and replayable observability.

A real smart home built like production infrastructure. Local-first, AI-powered, fully self-hosted: Home Assistant, Frigate, edge LLMs, agentic irrigation, and a 9U rack that runs the house.

The open Company State Runtime — version control for your project's state. Goals, decisions, work, and evidence outlive every chat, context window, and agent session. Ships the self CLI.

transformers safetensors mistral text-generation arxiv:2310.11511 arxiv:2306.02707

Knowledge retrieval (RAG) decorators for Azure Functions Python v2

transformers gguf text-generation zh arxiv:1910.09700 license:apache-2.0

把操作手机点外卖封装成给 AI agent 调用的工具——文本进文本出(markdown表格)、对非视觉/远程 LLM 友好、可 MCP 封装或原生 tool 直用。基于 Open-AutoGLM。

Bun-native CLI agent: long-term memory, knowledge graph, multi-channel (Telegram/Slack/Discord), 25+ LLM providers.

Build an agent in minutes. Add production controls when you need them.

Agent Lifecycle Framework: Documentation as executable infrastructure for birthing and evolving autonomous AI agents

Public reference operating model for reusable, evidence-driven AI agent teams on Multica.

Portable Agent Skills for disciplined software delivery, learning, editing, and LLM evaluation

Compare AI subscription plans, token pricing, and model costs across ChatGPT, Claude, Gemini, Cursor, OpenRouter, Kimi, and more.

Companion project for my course "AI Engineering Fundamentals - Build and Deploy real LLM apps with Python"

Local-first engine that grounds an AI agent's answers in your own Markdown and captures what it learns back into the KB — Node · TypeScript · MCP.

面向中小企业的开源私有化知识库：单实例部署、数据本地可控、LLM 中立。LangChain4j RAG、PgVector 混合检索、Docker 一键部署。

Open-source chat frontend + agent server for the Mastra AgentController (Agent Harness). Tool approvals, subagents, observational memory, schedules, and a live workspace — ~50 controller events wired to Vercel AI Elements. Clone the whole stack, or install either half with shadcn.

xCogito on YouTube

本象协议 Benxiang · 一源万影：保存本象，按需投影。面向 AI 的持久对象表示层——AI 只提交语义事务，确定性编译器校验后才落地，每个字段都能回答「凭什么」。A2 消融实测（2026-08-13，README §5）：唯一幸存的分类性主张是结构性的——提示词结构上产不出证据链；A3 全套机制给 15.0%/27.5% 受测字段可溯源证据，所有纯提示词基线均为 0%（A0→A2 +41.3pt 显著 p=3.3e-5，A2→A3 +2.5pt 不显著）。四方言共用核心仅加 4 行；87 条语言无关一致性向量 + 两份实现。

MUSS: Multilevel Subset Selection for relevance and diversity at scale (UAI 2026) — up to 80x faster than MMR with approximation guarantees, for RAG and candidate retrieval

AIpha Learning OS — 个性化学习多智能体系统，融合 Multi-Agent、RAG、课程知识构建流水线、Neo4j 知识图谱、学习记忆与任务级 LLM 路由。

ChatSafetyAI is the ChatGPT for field-based occupational safety

A RAG pipeline engine for scale

Agentic AI Engineer Learning Track - Build production multi-agent systems (agent loops, orchestration, RAG, eval, guardrails)

Production-grade agentic framework for Java. Simple, explicit, no magic.

MCP server for Yandex Wiki: full-text search, pages, comments, attachments and dynamic tables. 33 typed tools, server-side read-only mode, Docker-ready.

Laravel-native LLM observability, traces, cost tracking, latency, errors, and alerts for your AI calls

Your self-hosted personal AI agent — an agentic OS for one person: chat, autonomous goal loops, memory, knowledge base, skills, automation, MCP in both directions, and a permission-gated app platform. Local-first, provider-agnostic, MIT.

telegram AI that runs on your account. Answers in groups and PMs from personal account

Hermes Agent on Android — a community mobile port built with Python and Flet

6 ajanlı otonom multi-agent orkestrasyon sistemi — TypeScript, pgvector hafıza, öğrenen yönlendirme. Kendi sunucunda çalışır, BYOK (OpenRouter/OpenAI/Gemini/Ollama).

AstrBot plugin bridging DeepSeek Harness (DSH) agents — AstrBot 插件：桥接 DeepSeek Harness 智能体

DocsGPT is a powerful web app that allows you to embed your documents then query them using natural language